Faculty Bibliography

This paper proposes a new approach for functional analysis of cardiac images. A generic heart model is coupled with finite element methods to assess global and regional function from tagged magnetic resonance images. A model including both the left ventricle (LV) and right ventricle (RV) up to the basal area is needed for comprehensive understanding of cardiac physiology and anatomy. Most existing techniques can only handle the LV. Although some have attempted modeling the whole heart, these models have no geometric reference frame, so that it is difficult to compare different heart shapes and their motion. This paper uses a generic bi-ventricular heart model for functional analysis of heart motion. Three orthogonal tagging directions provide temporal correspondence of material points and enable tracking material points over time. The generic finite element model deforms due to forces exerted from material points by solving governing equations based on physical laws. The resulting model parameters can be used to characterize myocardial motion and deformation, including the basal area. We discuss the possibility of classifying parameters associated with normal and pathological hearts

Anatomical structures contain various types of curvilinear or tube-like structures such as blood vessels and bronchial trees. In medical imaging, the extraction and representation of such structures are of clinical importance. Complex curvilinear structures can be best represented by their center lines (or skeletons) along their elongated direction. In this paper, a gradient-based method for ridge point extraction on tubular objects is presented. Using the gradients of distance maps or intensity profiles usually generates skeleton surfaces for 3D objects, which is not desirable for representing tubular objects. To extract only the points on the centerline, we first employ the gradient vector flow (GVF) technique and then apply eigenanalysis of the Hessian matrix to remove false positive points. We present various results of the method using CLSM (Confocal Laser Scanning Microscopy) images of blood fibrins and CT images of a skull and lungs. Our method is efficient and allows for completely automatic extraction of points along the centerline of a tubular object in its elongated direction

BACKGROUND: Magnetic resonance (MR) imaging is frequently used to diagnose arrhythmogenic right ventricular cardiomyopathy/dysplasia (ARVC/D). However, the reliability of various MR imaging features for diagnosing ARVC/D is unknown. The purpose of this study was to determine which morphologic MR imaging features have the greatest interobserver reliability for diagnosing ARVC/D. METHODS: Forty-five sets of films of cardiac MR images were sent to 8 radiologists and 5 cardiologists with experience in this field. There were 7 cases of definite ARVC/D as defined by the Task Force criteria. Six cases were controls. The remaining 32 cases had MR imaging because of clinical suspicion of ARVC/D. Readers evaluated the images for the presence of (a) right ventricle (RV) enlargement, (b) RV abnormal morphology, (c) left ventricle enlargement, (d) presence of high T(1) signal (fat) in the myocardium, and (e) location of high T(1) signal (fat) on a Likert scale with formatted responses. RESULTS: Readers indicated that the Task Force ARVC/D cases had significantly more (chi(2) = 119.93, d.f. = 10, p < 0.0001) RV chamber size enlargement (58%) than either the suspected ARVC/D (12%) or no ARVC/D (14%) cases. When readers reported the RV chamber size as enlarged they were significantly more likely to report the case as ARVC/D present (chi(2)(= )33.98, d.f. = 1, p < 0.0001). When readers reported the morphology as abnormal they were more likely to diagnose the case as ARVC/D present (chi(2) = 78.4, d.f. = 1, p < 0.0001), and the Task Force ARVC/D (47%) cases received significantly more abnormal reports than either suspected ARVC/D (20%) or non-ARVC/D (15%) cases. There was no significant difference between patient groups in the reported presence of high signal intensity (fat) in the RV (chi(2) = 0.9, d.f. = 2, p > 0.05). CONCLUSIONS: Reviewers found that the size and shape of abnormalities in the RV are key MR imaging discriminates of ARVD. Subsequent protocol development and multicenter trials need to address these parameters. Essential steps in improving accuracy and reducing variability include a standardized acquisition protocol and standardized analysis with dynamic cine review of regional RV function and quantification of RV and left ventricle volumes

We describe an automated, model-based method to segment the left and right ventricles in 4D tagged MR. We fit 3D epicardial and endocardial surface models to ventricle features we extract from the image data. Excellent segmentation is achieved using novel methods that (1) initialize the models and (2) that compute 3D model forces from 2D tagged MR images. The 3D forces guide the models to patient-specific anatomy while the fit is regularized via internal deformation strain energy of a thin plate. Deformation continues until the forces equilibrate or vanish. Validation of the segmentations is performed quantitatively and qualitatively on normal and diseased subjects.

Little information is known about in-vivo heart strain and stress distribution. In this paper, we present a novel statistical model to estimate. the in-vivo material properties and strain and stress distribution in the left ventricle. The displacements of the heart wall are reconstructed in previous work of our group by using MRI-SPAMM tagging technique and deformable model. Based on the reconstructed displacements, we developed the statistical model to estimate strain and stress by using EM algorithm. Two normal hearts and two hearts with right-ventricular hypertrophy are studied. We find noticeable differences in the strain and stress estimated for normal and abnormal hearts

Making a generic heart deformable model to be able to analyze normal and pathological hearts is important. Such a generic model gives more stability and accuracy for segmentation, analysis and classsification. Due to the conflicting demands of shape generality and shape compactness, such a generic heart model is difficult to define. In order to be useful, a generic heart model should be defined with a few number of parameters. In all the previous work on the modeling of the LV-RV shape the deformable model is built from the given datasets. Therefore such methods have limitations that the quality of shape estimation are dependent on the quality of the datasets. In this paper, we introduce a blended deformable model approach with parameter functions which is generic enough to deal with the different heart shapes. Using a method we are able to model the 3D shape of the heart which include the left ventricle(LV) and the right ventricle(RV). We also include the inflow and outflow tract of the RV basal area, so that the full LV-RV shape can be estimated

In this paper we describe a completely automated volume-based method for the segmentation of the left and right ventricles in 4D tagged MR (SPAMM) images for quantitative cardiac analysis. We correct the background intensity variation in each volume caused by surface coils using a new scale-based fuzzy connectedness procedure. We apply 3D grayscale opening to the corrected data to create volumes containing only the blood filled regions. We threshold the volumes by minimizing region variance or by an adaptive statistical thresholding method. We isolate the ventricular blood filled regions using a novel approach based on spatial and temporal shape similarity. We use these regions to define the endocardium contours and use them to initialize an active contour that locates the epicardium through the gradient vector flow of an edgemap of a grayscale-closed image. Both quantitative and qualitative results on normal and diseased patients are presented.

PURPOSE: To investigate the use of a three-dimensional rapid acquisition with relaxation enhancement (RARE) pulse sequence for direct acquisition of phosphocreatine (PCr) images of the human myocardium. MATERIALS AND METHODS: A short elliptical birdcage radiofrequency (RF) body coil was constructed to produce a uniform flip angle throughout the chest cavity. In vivo images using a spectrally-selective RARE sequence with a spatial resolution of 1.2 cm x 1.2 cm x 2.5 cm (4 cm(3)) were acquired in nine minutes and 40 seconds. RESULTS: Scans of phantoms demonstrated excellent spectral selectivity. The signal-to-noise ratio in the myocardium ranged from 12.6 in the anterior wall to 5.3 in the mid septum. CONCLUSION: This study demonstrates that PCr data can be acquired using a three-dimensional RARE sequence with greater spatial and temporal resolution than spectroscopic techniques

Magnetic resonance imaging (MRI) provides a noninvasive way to evaluate the biomechanical dynamics of the heart. MRI can provide spatially registered tomographic images of the heart in different phases of the cardiac cycle, which can be used to assess global cardiac function and regional endocardial surface motion. In addition, MRI can provide detailed information on the patterns of motion within the heart wall, permitting calculation of the evolution of regional strain and related motion variables within the wall. These show consistent patterns of spatial and temporal variation in normal subjects, which are affected by alterations of function due to disease. Although still an evolving technique, MRI shows promise as a new method for research and clinical evaluation of cardiac dynamics

OBJECTIVES: This study was designed to validate strain measurements obtained using magnetic resonance tagging with spatial modulation of magnetization (SPAMM). We compared circumferential segment shortening measurements (%S) obtained using SPAMM to sonomicrometry %S in a canine model with (n = 28) and without (n = 3) coronary artery ligation. BACKGROUND: Magnetic resonance tagging enables noninvasive measurement of myocardial strain, but such strain measurements have not yet been validated in vivo. METHODS: Circumferential sonomicrometry crystal pairs were placed in apical myocardium at ischemic risk in ligation studies and in adjacent and remote myocardium. The %S was obtained from closely juxtaposed sonomicrometry and SPAMM sites. RESULTS: Paired data were available from 19 of 31 studies. Both methods distinguished remote from ischemic function effectively (p = 0.014 for SPAMM and p = 0.002 for sonomicrometry). SPAMM %S was similar to sonomicrometry %S in ischemic myocardium (2 +/- 3 vs. 0 +/- 3 p = 0.067) but was slightly higher than sonomicrometry %S in remote myocardium (11 +/- 10 vs. 7 +/- 5, p = 0.033). End-systolic (n = 30) and late systolic (n = 34) SPAMM %S correlated well with sonomicrometry %S (r = 0.84, p < 0.0001 and r = 0.88, p < 0.0001). CONCLUSIONS: Magnetic resonance tagging using SPAMM can quantitate myocardial strain in ischemic and remote myocardium. This study validates its application in scientific investigation and clinical assessment of patients with myocardial ischemia